about RandomCrop

[babyta]

Hello, great work, I would like to ask, after the training mask is randomly cropped, how to ensure that it is consistent with the original image?

[skuley]

Hello @babyta

I am afraid that i didn't fully understand your question. Why would you want randomly cropped train dataset to be consistent to original image?

If I understood correctly, you randomly crop train image by p-value(with probability), thus original image can used too without any random crop.

Hope this answered your question